Comparison of Frailty Scores in Newly Diagnosed Patients with Multiple Myeloma: A Review.

Introduction

Multiple myeloma is a neoplasm characterized by the clonal proliferation of malignant plasma cells within the bone marrow ( 1 ). It accounts for approximately 13% of all hematological malignancies, and 20% of hematological malignancy-related deaths ( 2 , 3 ). In Western countries, the median age at diagnosis is 70 years, making myeloma a disease burden for older adults, with 35 – 40% of patients being older than 75 years ( 4 ).

The introduction of novel therapeutic agents (immunomodulatory drugs, proteasome inhibitors, and monoclonal antibodies) and better supportive care have improved outcomes in patients with myeloma, with a substantial benefit in both progression-free survival (PFS) and overall survival (OS) ( 5 ). Although there has been improvement in outcomes for older adults, several clinical trials and population-based registries demonstrate that this benefit has been less pronounced in the ‘oldest' subgroup of patients age 75 or older ( 2 , 6 , 7 ), though that gap may be closing ( 8 ). The challenge in treating older patients is to accurately identify ‘fit' patients who can tolerate more intensive regimens and potentially a stem cell transplant, while simultaneously identifying ‘frail' patients who are transplant ineligible and require substantial modification of the regimen and dose of individual drugs to decrease morbidity and mortality.

A geriatric assessment is one tool which has been utilized in the oncologic setting to risk stratify the ‘frailty' of patients. It includes assessment of comorbidities, functional status, medications (including polypharmacy and inappropriate medication), cognition, psychological status, and social support [ 9 ]. Over the past two decades, a large body of literature has developed, demonstrating that geriatric assessment provides a better picture of the health of older adults than traditional oncologic assessments of performance status ( 10 , 11 ). A geriatric assessment is also known to impact treatment decision in older patients with cancer in up to 2/3 of the patients, with most recommendations being attenuated therapy in more frail patients ( 12 ).

There is no clear consensus of frailty in the setting of multiple myeloma; however, multiple tools have been developed to help identify ‘frail' patients and stratify older patients with myeloma. The International Myeloma Working Group (IMWG) Frailty Score is based on chronological age plus 3 tools: Katz Activity of Daily Living, the Lawton Instrumental activity of Daily Living and the Charlson Comorbidity Index. Patients categorized as frail are more likely to experience grades 3–4 non-hematologic toxicity of therapy, early discontinuation of treatment and a shorter OS ( 13 ). The Mayo Clinic Frailty Score is a tool based upon age, Eastern Cooperative Oncology Performance Status (ECOG PS), and the serum biomarker N-terminal fragment of the type-B natriuretic peptide (NT-proBNP) level ( 14 ). Another tool, the Revised Myeloma Comorbidity Index (R-MCI) is also associated with OS and accounts for several prognostic factors including chronological age, impaired lung and kidney function, the karnofsky performance status, frailty, and unfavourable cytogenetics ( 15 ). While the IMWG and R-MCI scores rely more on the ‘fitness' level of a patient as judged by their ability to function in different domains of daily living, the Mayo Clinic Frailty Score incorporates biomarkers in defining frailty as well.

In this review, we will examine and critically appraise three commonly utilized frailty scores in myeloma (IMWG Frailty Score, Mayo Clinic Frailty Score, R-MCI Index). Recent reviews have summarized these scores ( 16 , 17 ); our aim was to extend these reviews and explore in greater depth the derivations of these scores focusing on the patient population, methods and validations of these tools in clinical setting.

Derivation of the Frailty Scores

Patient characteristics of the prognostic scores

In order to understand the applicability of the frailty prognostic scores in myeloma, it is essential to understand the population base that was used to develop these scores. The patient characteristics used to develop each prognostic score are detailed below and a summary is provided in Table 1 .

IMWG Frailty Score

Palumbo et al. created a score using data from patients with newly-diagnosed multiple myeloma that were enrolled in three prospective multicenter clinical trials ( 13 ). The population inclusion criteria included all ages; however, patients had to be ineligible for transplant. Most of the patients were recruited from Italy and the Czech Republic. A total of 869 patients were included in the analysis with a median age of 74 years. Given that the patients were recruited from clinical trials, despite their ‘older' age, they were relatively fit, with only 21% of the patients with an ECOG PS status of 2 or greater. Most of the patients in the cohort received novel agents and about one-third had higher stage disease (International Staging System [ISS] Stage III).

Mayo Clinic Frailty Score

Milani et al. created a score using data from 351 patients with newly-diagnosed multiple myeloma that were enrolled in a prospective registry at a single institution from 2007–2011 ( 14 ). As both transplant eligible and ineligible patients were included in the analysis, the median age was lower than that of the IMWG, at 65 years. Comparable to IMWG, approximately one-fifth of the patients had an ECOG PS of 2 or greater, and one-third has an ISS stage of III. Although greater than 63% of the patients received novel agents as part of the treatment, only 39% underwent transplantation.

Revised Myeloma Co-Morbidity Index

Engelhardt et al. created a prognostic score prospectively consisting of all newly-diagnosed myeloma patients that presented consecutively to their institution ( 15 ). Their population was younger, with a median age of 63; however, as all patients were included, there was a wide age distribution from 21 to 93 years. Overall, there were slightly more patients that had higher stage disease (41% with ISS Stage III) compared to the other two scores. Approximately half of the patients were treated with transplant and at least 63% had novel agents incorporated into their therapy.

Methodology in development of the prognostic score

Rigorous development of a prognostic model is essential in its ability to provide useful information regarding patient prognosis and potentially for informing patient treatment. The development of each frailty index is outlined below and the key characteristics are summarized in Table 2 .

IMWG Frailty Score

For the development of this frailty index, patients from three clinical trials were pooled. Patients were analyzed on an intention-to-treat basis in the original clinical trials used as a dataset for this score. Multiple confounding factors were considered that are clinically relevant, including the staging (ISS). Although the newer revised international staging system ( 18 ), which is currently in use, was not accounted for given the time frame of the study, other high-risk features such as cytogenetics were included. The final model developed using an additive frailty score based upon the integer part of the hazard ratio for each prognostic factor that was significant in the multivariate analysis. The score system (range 0–5), based on age, comorbidities (Charlson Comorbidity Index), and physical function (Katz Activity of Daily Living and Lawton Instrumental Activity of Daily Living) identified 3 groups of patients: fit (score=0), intermediate-fitness (score=1), and frail (score≥2).

In their final model, the 3-year overall survival was 84% in fit patients, 76% in intermediate-fitness patients (Hazard Ratio [HR] 1.61, 95% Confidence Interval [CI] 1.02–2.56, p=0.042) and 57% in frail patients (HR 3.57 CI 95% 2.37-5.39, p<0.001). The authors also used their model to predict additional outcomes such as the cumulative incidence of grade ≥3 nonhematologic adverse events and treatment discontinuation at 12 months in each risk group.

Mayo Clinic Frailty Score

The cornerstone of this model was the measurement of the biomarker NT-proBNP that was conducted using one specific assay on frozen sera sample. While the authors do acknowledge the relationship NT-proBNP with age, they do not show account for this in their model and have rather created a stringed dichotomized variable. NT-proBNP is also affected by decreased renal function; however, it is not entirely clear how the prognostic ability of this model changes for those subset of myeloma patients presenting with renal dysfunction. The final proposed model is the ‘simplest' model consisting of age ≥70, ECOG PS ≥2, and the serum biomarker NT-proBNP level of ≥300 ng/L which a score of 1 for each of these variables. This final model allowed patients to be categorized into four risk groups with scores of 0-3 point respectively and is associated with a median OS not reached with 0 risk factors, 58 months with one risk factor, 28 months with two risk factors and 18 months with all of the three risk factors.

Revised Myeloma Co-Morbidity Index

For the development of the prognostic score, 552 patients were included. Based upon the authors' previous studies and analyses, which led to the initial additive Myeloma Co-morbidity Index, impaired renal function, lung function and karnofsky performance status were known to be relevant risk factors for OS and PFS ( 19 , 20 , 21 ). A multivariate analysis with backward selection for OS to evaluate the significance of the additional variables was conducted. The model identified a total of 5 variables that were associated with OS including age, renal status, lung dysfunction, Karnofsky Performance Status and frailty status (includes Karnofsky Performance Status < 70%, timed up and go test < 10 sec, Instrumental Activity of Daily Living score ≤ 4, and subjective fitness). Cytogenetics could also be added if available; however, were not essential in the calculation of the score. A score was assigned to each of the five predicted variables based upon the log hazard ratio leading to a cumulative score of 9 points. The patients were characterized into ‘fit', ‘intermediate-fit' and ‘frail' with median OS of 10.1, 4.4 and 1.2 years.

Validation of the prognostic scores

Validation of a prognostic model is essential in establishing whether the model is generalizable to other similar populations than the ones used in the original dataset ( 22 ). A model can be validated either internally or externally. External validation is considered more stringent as it tests the generalizability of the model in an entirely different cohort and is considered essential prior to the broad usage of a prognostic score in daily clinical practice ( 23 ). The validation for each frailty index is outlined below.

IMWG Frailty Score

No internal validation cohort was conducted as part of the original study by Palumbo et al ( 13 ). As the authors acknowledge themselves, this is a major limitation of their study. They chose not to split their sample into a derivation and validation cohort as they felt their sample size was inadequate. External validation of the IMWG score was done by prospectively analyzing 125 patients at a single institution in Germany ( 24 ). While this prospective cohort and the original IMWG cohort were similarly matched for most baseline characteristics, the validation cohort had a younger population, with over 59% under age 65, compared to 2% in the IMWG cohort. As this was a real-world population, there was a higher percentage of patients identified as frail (48% versus 30% in the IMWG cohort) despite the younger age range of patients in the validation cohort. The IMWG was prognostic for OS and PFS in this validation cohort; however, both the OS and PFS were shorter in the validation cohort than the derivation cohort for the intermediate-fit and frail patients. This external validation of the IMWG score adds to the validity of the score and furthermore suggests its prognostic value, even in transplant eligible patients, which were originally not included in the derivation sample. To our knowledge, there has been no further updating or adjustment of the IMWG frailty index following this external validation.

Mayo Clinic Frailty Score

No validation cohort data was presented as part of the original derivation and to date, there is no published external validation cohort.

Revised Myeloma Co-Morbidity Index

The R-MCI was internally validated; the cohort of 810 patients was randomly split into 2/3 and 1/3, training (n=552) and validation set (n=249). Since that time, there have been additional studies demonstrating the treatment responsiveness of R-MCI in different clinical setting. Not only are the proportion of patients categorized as fit or frail found to be comparable in real world setting as those on clinical trials ( 25 ), additional studies show that R-MCI along with the IMWG as a longitudinal measure is responsive to change and the score may improve in patients responding to different types of myeloma therapy ( 26 , 27 ).

Usability and applicability of the prognostic score

Studies have shown that although a large number of prognostic scores are published, very few models are actually used in clinic practice ( 28 ). Although this is likely due to a number of factors, the ease of utility and the practicality of the score is essential for it to enter into routine clinical practice. There needs to be sufficient reporting that is presented in an easily accessible way regarding how to compute a risk score and also a link to the corresponding survival probabilities. Currently, there is no data published on the interrater reliability on these three scores. Further information regarding the usability of each score is outline below:

IMWG Frailty Score

The score is currently available as an online tool at http://www.myelomafrailtyscorecalculator.net/ . There are 31 categories that need to be filled out; however, as all the variables are self-reported by the patient, no additional investigations are required improving its usability in a busy clinic schedule. The online tool only requires us to fill out categories with deficits. The average time to calculate this score is estimated to be between 5 to 7 minutes ( 13 ).

Mayo Clinic Frailty Score

This prognostic score is easy to use given that it only consists of three variables. Age and ECOG-PS are routinely measured by oncologists. The NT pro-BNP level is a blood test which is available in most laboratories; however, the sensitivity and specificity of the assay differs across different labs and may be variable. Additionally, this test is not routinely recommended as part of the evaluation of myeloma [ 29 , 30 ], has cost associated, depending upon resource constraints, it may not be available to all demographics. When the NT-proBNP level is available, the score can likely be calculated in 1–2 minutes.

Revised Myeloma Co-Morbidity Index

In the derivation of the score, the assessment of each variable in this index was performed by staff members with specialized training in oncogeriatrics. This score is currently available online http://www.myelomacomorbidityindex.org/en_calc.html . It has an easy drop-down menu format with the definition of each variable listed below. The variables included in this score include additional investigations such as pulmonary function tests. Additionally, the variables used to calculate frailty consist of multiple objective and subjective tools such as timed up and go test and fitness respectively which oncologists do not routinely utilize; however are overall simple tests to conduct. The score can be calculated on the website and the output is an easily interpretable Kaplan-Meier survival curve ( 24 ).

Discussion

The three frailty scores created for multiple myeloma have provided clinicians with additional information regarding prognostication of their patients. There was a wide range in the applicability of these frailty measurements: from short, fast to more time-consuming and sophisticated measurements of frailty. Clinicians will have to be aware that each of these scores was tested in a specific patient population and derived during a specific time period. Furthermore, as new treatment modalities and new drugs are introduced into the treatment algorithm, the natural history of myeloma may change and therefore these prognostic scores will need to be updated and validated continuously.

Although most clinicians believe that routine frailty assessment is time-consuming and costly, research has shown that, in fact, not only is frailty assessment feasible, it is also likely less time-consuming and costly than many other prognostic tests done in oncology ( 31 ). Targeting the most promising tools and building the evidence-base in prospective studies to support and establish their psychometric and clinical-metric features has been an important research priority for the past decade and will require ongoing collaboration between groups. Specific tools, for example, may be better for population-level frailty screening for patients with newly-diagnosed myeloma; others may be best suited for a comprehensive assessment to predict toxicity of specific treatment regimens. The role of biomarkers, such as the senescence marker p16INK4a and sarcopenia, in addition to objectively measured functional geriatric assessments, such as gait speed and handgrip strength, may provide further opportunities for refinement of frailty indices ( 17 , 32 ).

Although none of these tools has been derived in patients with relapsed or refractory myeloma, nor evaluated longitudinally, they may be applicable in this patient cohort as well and may represent an additional tool to understand treatment outcomes in patients on clinical trials or undergoing different treatment modalities ( 26 , 27 , 33 ). This represents an important area for further research, as the concept of ‘frailty' may be even more important in relapsed patients who have more aggressive disease, accumulated more co-morbidities ( 34 ) and who have residual toxicities from previous myeloma treatments. The role of these scores in understanding the experiences of patients undergoing transplantation and in other plasma cell dyscrasias will also need to be further explored.

Based upon the multiple tools, the IMWG frailty score and the R-MCI index appears to be the most comprehensive tools for risk stratification. A limitation of all those scores, including the IMWG, is the inclusion of chronological age, which automatically increases ‘frailty' without taking biological or functional age ( 35 ) into account and may not accurately represent outcomes for functionally vigorous older individuals ( 36 , 37 ).

After defining a specific frailty measurement, the next step in improving the usability and the clinical utility of these scores will be to conduct studies aimed at certain interventions for each risk groups, which can lead to a potential change in health outcomes. At a public health level, for example, if a patient is identified as frail, perhaps additional home resources or early initiation of palliative care may be justified. At a physician level, understanding the overall prognosis of a patient may lead to specific interventions such as initial dose reduction or alternative chemotherapy regimens. Several groups have proposed dose modifications for ‘frail' vs ‘fit' patients based upon the IMWG and R-MCI frailty scores ( 16 , 17 , 38 ). While they represent an important step in the operationalization of frailty in routine clinical practise, the prospective demonstration of the impact of frailty-based treatment arms in clinical trials is now increasingly being recognized ( 39 , 40 ). The results of the phase III clinical trial designed for intermediate fit (as defined by the IMWG Frailty Score) in older patients with newly diagnosed multiple myeloma were recently presented highlighting similar efficacy and improved tolerability of a dose reduced regimen of Lenalidomide and Dexamethasone as compared to standard continuous therapy ( 39 ). Additionally, a randomized controlled trial is currently being conducted with over 700 older patients with newly diagnosed myeloma being assigned to either a standard arm or a frailty adapted regimen based upon the IMWG Frailty Score (Myeloma XIV: Frailty-adjusted Therapy in Transplant Non-Eligible Patients with Newly Diagnosed Multiple Myeloma, NCT03720041 ). These recent studies highlight the exciting advancement that is occurring in understanding the prognostic implication of frailty in multiple myeloma.

Ongoing areas of research include the associations between frailty and patient-centered outcomes as they may in fact be most relevant to patients. The IMWG Frailty Scores is associated with toxicity and treatment discontinuation, which are more patient-centered, though direct patient-reported outcomes such as quality-of-life have not been examined. Other geriatric outcomes such as future risk of falls and the effect of quality of life measurements will also need to be incorporated into further development and validation of any myeloma specific prognostic scores.

Lastly, from a methodological perspective, whereas in the past, the credibility of prognostic scores was mostly derived from individual studies, recently there has been research and development into using systematic reviews for understanding the comprehensive body of evidence and a growing number of systematic reviews in this field are being published ( 41 ). Furthermore, Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach is increasingly being optimized order to understand the strength of inferences from the combined summarization of these individual prognostic studies ( 42 ). Currently, there exists no formal systematic review of the common frailty scores in myeloma; however, that may represent an important future direction in trying to understand the overall assessment of the prognostic estimates from these individual studies.

Funding: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; in the preparation of the manuscript; or in the review or approval of the manuscript.

Conflicts of Interest: The authors have no conflict of interest pertaining to the subject matter in the manuscript.

References

1. Swerdlow S C, Harris NL, Jaffe ES, Pileri SA, Stein H, et al. IARC Press; France: 2008. Lyon, France.

2. Pozzi S, Marcheselli L, Bari A, Liardo EV, Marcheselli R, Luminari S, et al. Survival of multiple myeloma patients in the era of novel therapies confirms the improvement in patients younger than 75 years: a population-based analysis. British journal of haematology. 2013;163(1):40–46. doi: 10.1111/bjh.12465. 10.1111/bjh.12465

3. Palumbo A, Anderson K. Multiple myeloma. The New England journal of medicine. 2011;364(11):1046–1060. doi: 10.1056/NEJMra1011442. 10.1056/NEJMra1011442

4. Palumbo A, Bringhen S, Ludwig H, Dimopoulos MA, Blade J, Mateos MV, et al. Personalized therapy in multiple myeloma according to patient age and vulnerability: a report of the European Myeloma Network (EMN) Blood. 2011;118(17):4519–4529. doi: 10.1182/blood-2011-06-358812. 10.1182/blood-2011-06-358812

5. Kumar SK, Rajkumar SV, Dispenzieri A, Lacy MQ, Hayman SR, Buadi FK, et al. Improved survival in multiple myeloma and the impact of novel therapies. Blood. 2008;111(5):2516–2520. doi: 10.1182/blood-2007-10-116129. 10.1182/blood-2007-10-116129

6. Hulin C, Belch A, Shustik C, Petrucci MT, Duhrsen U, Lu J, et al. Updated Outcomes and Impact of Age With Lenalidomide and Low-Dose Dexamethasone or Melphalan, Prednisone, and Thalidomide in the Randomized, Phase III FIRST Trial. J Clin Oncol. 2016;34(30):3609–3617. doi: 10.1200/JCO.2016.66.7295. 10.1200/JCO.2016.66.7295

7. Mateos MV, Oriol A, Martinez-Lopez J, Gutierrez N, Teruel AI, de Paz R, et al. Bortezomib, melphalan, and prednisone versus bortezomib, thalidomide, and prednisone as induction therapy followed by maintenance treatment with bortezomib and thalidomide versus bortezomib and prednisone in elderly patients with untreated multiple myeloma: a randomised trial. Lancet Oncol. 2010;11(10):934–941. doi: 10.1016/S1470-2045(10)70187-X. 10.1016/S1470-2045(10)70187-X

8. Kumar SK, Dispenzieri A, Lacy MQ, Gertz MA, Buadi FK, Pandey S, et al. Continued improvement in survival in multiple myeloma: changes in early mortality and outcomes in older patients. Leukemia. 2014;28(5):1122–1128. doi: 10.1038/leu.2013.313. 10.1038/leu.2013.313

9. Wildiers H, Heeren P, Puts M, Topinkova E, Janssen-Heijnen ML, Extermann M, et al. International Society of Geriatric Oncology consensus on geriatric assessment in older patients with cancer. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2014;32(24):2595–2603. doi: 10.1200/JCO.2013.54.8347. 10.1200/JCO.2013.54.8347

10. Extermann M, Overcash J, Lyman GH, Parr J, Balducci L. Comorbidity and functional status are independent in older cancer patients. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 1998;16(4):1582–1587. doi: 10.1200/JCO.1998.16.4.1582. 10.1200/JCO.1998.16.4.1582

11. Repetto L, Fratino L, Audisio RA, Venturino A, Gianni W, Vercelli M, et al. Comprehensive geriatric assessment adds information to Eastern Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2002;20(2):494–502. doi: 10.1200/JCO.2002.20.2.494. 10.1200/JCO.2002.20.2.494

12. Hamaker ME, Schiphorst AH, ten Bokkel H D, Schaar C, van Munster BC. The effect of a geriatric evaluation on treatment decisions for older cancer patients—a systematic review. Acta Oncol. 2014;53(3):289–296. doi: 10.3109/0284186X.2013.840741. 10.3109/0284186X.2013.840741

13. Palumbo A, Bringhen S, Mateos MV, Larocca A, Facon T, Kumar SK, et al. Geriatric assessment predicts survival and toxicities in elderly myeloma patients: an International Myeloma Working Group report. Blood. 2015;125(13):2068–2074. doi: 10.1182/blood-2014-12-615187. 10.1182/blood-2014-12-615187

14. Milani P, Vincent R S, Merlini G, Kumar S, Gertz MA, Palladini G, et al. N-terminal fragment of the type-B natriuretic peptide (NT-proBNP) contributes to a simple new frailty score in patients with newly diagnosed multiple myeloma. American journal of hematology. 2016;91(11):1129–1134. doi: 10.1002/ajh.24532. 10.1002/ajh.24532

15. Engelhardt M, Domm AS, Dold SM, Ihorst G, Reinhardt H, Zober A, et al. A concise revised Myeloma Comorbidity Index as a valid prognostic instrument in a large cohort of 801 multiple myeloma patients. Haematologica. 2017;102(5):910–921. doi: 10.3324/haematol.2016.162693. 10.3324/haematol.2016.162693

16. Larocca A, Dold SM, Zweegman S, Terpos E, Wasch R, D'Agostino M, et al. Patient-centered practice in elderly myeloma patients: an overview and consensus from the European Myeloma Network (EMN) Leukemia. 2018;32(8):1697–1712. doi: 10.1038/s41375-018-0142-9. 10.1038/s41375-018-0142-9

17. Zweegman S, Engelhardt M, Larocca A. ‘Aging ESo, Hematology. Elderly patients with multiple myeloma: towards a frailty approach? Current opinion in oncology. 2017;29(5):315–321. doi: 10.1097/CCO.0000000000000395. 10.1097/CCO.0000000000000395

18. Palumbo A, Avet-Loiseau H, Oliva S, Lokhorst HM, Goldschmidt H, Rosinol L, et al. Revised International Staging System for Multiple Myeloma: A Report From International Myeloma Working Group. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2015;33(26):2863–2869. doi: 10.1200/JCO.2015.61.2267. 10.1200/JCO.2015.61.2267

19. Kleber M, Ihorst G, Gross B, Koch B, Reinhardt H, Wasch R, et al. Validation of the Freiburg Comorbidity Index in 466 multiple myeloma patients and combination with the international staging system are highly predictive for outcome. Clinical lymphoma, myeloma & leukemia. 2013;13(5):541–551. doi: 10.1016/j.clml.2013.03.013. 10.1016/j.clml.2013.03.013

20. Kleber M, Ihorst G, Terhorst M, Koch B, Deschler B, Wasch R, et al. Comorbidity as a prognostic variable in multiple myeloma: comparative evaluation of common comorbidity scores and use of a novel MM-comorbidity score. Blood cancer journal. 2011;1(9):e35. doi: 10.1038/bcj.2011.34. 10.1038/bcj.2011.34

21. Kleber M, Ihorst G, Udi J, Koch B, Wasch R, Engelhardt M. Prognostic risk factor evaluation in patients with relapsed or refractory multiple myeloma receiving lenalidomide treatment: analysis of renal function by eGFR and of additional comorbidities by comorbidity appraisal. Clinical lymphoma, myeloma & leukemia. 2012;12(1):38–48. doi: 10.1016/j.clml.2011.09.216. 10.1016/j.clml.2011.09.216

22. Altman DG, Royston P. What do we mean by validating a prognostic model? Statistics in medicine. 2000;19(4):453–473. doi: 10.1002/(sici)1097-0258(20000229)19:4<453::aid-sim350>3.0.co;2-5. 10.1002/(SICI)1097-0258(20000229)19:4<453::AID-SIM350>3.0.CO;2-5

23. Bleeker SE, Moll HA, Steyerberg EW, Donders AR, Derksen-Lubsen G, Grobbee DE, et al. External validation is necessary in prediction research: a clinical example. Journal of clinical epidemiology. 2003;56(9):826–832. doi: 10.1016/s0895-4356(03)00207-5. 10.1016/S0895-4356(03)00207-5

24. Engelhardt M, Dold SM, Ihorst G, Zober A, Moller M, Reinhardt H, et al. Geriatric assessment in multiple myeloma patients: validation of the International Myeloma Working Group (IMWG) score and comparison with other common comorbidity scores. Haematologica. 2016;101(9):1110–1119. doi: 10.3324/haematol.2016.148189. 10.3324/haematol.2016.148189

25. Gengenbach L, Reinhardt H, Ihorst G, Ajayi S, Dold SM, Kohler M, et al. Navigating the changing multiple myeloma treatment landscape: clinical practice patterns of MM patients treated in- and outside German DSMM study group trials. Leukemia & lymphoma. 2018;23:1–8. doi: 10.1080/10428194.2018.1448084.

26. Greil C, Engelhardt M, Ihorst G, Schoeller K, Bertz H, Marks R, et al. Allogeneic transplantation of multiple myeloma patients may allow long-term survival in carefully selected patients with acceptable toxicity and preserved quality of life. Haematologica. 2018 doi: 10.3324/haematol.2018.200881.

27. Waldschmidt JM, Keller A, Ihorst G, Grishina O, Muller S, Wider D, et al. Safety and efficacy of vorinostat, bortezomib, doxorubicin and dexamethasone in a phase I/II study for relapsed or refractory multiple myeloma (VERUMM study: vorinostat in elderly, relapsed and unfit multiple myeloma) Haematologica. 2018;103(10):e473–e4e9. doi: 10.3324/haematol.2018.189969. 10.3324/haematol.2018.189969

28. Steyerberg EW. Springer; New York: 2008. Clinical Prediction Models.

29. Rajkumar SV, Dimopoulos MA, Palumbo A, Blade J, Merlini G, Mateos MV, et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. The Lancet Oncology. 2014;15(12):e538–e548. doi: 10.1016/S1470-2045(14)70442-5. 10.1016/S1470-2045(14)70442-5

30. National Comprehensive Cancer Network Myeloma Guidelines. https://www.nccnorg/professionals/physician_gls/defaultaspx

31. Hamaker ME, Wildes TM, Rostoft S. Time to Stop Saying Geriatric Assessment Is Too Time Consuming. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2017;35(25):2871–2874. doi: 10.1200/JCO.2017.72.8170. 10.1200/JCO.2017.72.8170

32. Rosko AE, Huang Y, Benson DM, Efebera YA, Hofmeister C, Jaglowski S, et al. Use of a comprehensive frailty assessment to predict morbidity in patients with multiple myeloma undergoing transplant. Journal of geriatric oncology. 2018 doi: 10.1016/j.jgo.2018.05.015.

33. Facon T, Hulin C, Dimopoulos MA, Belch A, Meuleman N, Mohty M, et al. A Frailty Scale Predicts Outcomes of Patients with Newly Diagnosed Multiple Myeloma Who Are Ineligible for Transplant Treated with Continuous Lenalidomide Plus Low-Dose Dexamethasone on the First Trial. Blood. 2015;126(23):4239.

34. Hari P, Romanus D, Luptakova K, Blazer M, Yong C, Raju A, et al. The impact of age and comorbidities on practice patterns and outcomes in patients with relapsed/refractory multiple myeloma in the era of novel therapies. Journal of geriatric oncology. 2018;9(2):138–144. doi: 10.1016/j.jgo.2017.09.007. 10.1016/j.jgo.2017.09.007

35. Gay F, Engelhardt M, Terpos E, Wasch R, Giaccone L, Auner HW, et al. From transplant to novel cellular therapies in multiple myeloma: European Myeloma Network guidelines and future perspectives. Haematologica. 2018 doi: 10.3324/haematol.2017.174573.

36. Murillo A, Cronin AM, Laubach JP, Hshieh TT, Tanasijevic AM, Richardson PG, et al. Performance of the International Myeloma Working Group myeloma frailty score among patients 75 and older. Journal of geriatric oncology. 2018 doi: 10.1016/j.jgo.2018.10.010.

37. Mitnitski AB, Mogilner AJ, Rockwood K. Accumulation of deficits as a proxy measure of aging. ScientificWorldJournal. 2001;1:323–336. doi: 10.1100/tsw.2001.58. 10.1100/tsw.2001.58

38. Zweegman S, Palumbo A, Bringhen S, Sonneveld P. Age and aging in blood disorders: multiple myeloma. Haematologica. 2014;99(7):1133–1137. doi: 10.3324/haematol.2014.110296. 10.3324/haematol.2014.110296

39. Alessandra L M, De Lorenzo P, Nicola C, Giulia B, Monica G, Vittorio M, Tommaso C d T, Anna B, Stefano S, Nicola G, Chiara P, Stefano P, Sonia R, Norbert P, Anna M L, Francesca P, Claudia C, Patrizia T, Massimo O, Michele C, Antonio P, Mario B, Sara Bringhen. Efficacy and Feasibility of Dose/Schedule-Adjusted Rd-R Vs. Continuous Rd in Elderly and Intermediate-Fit Newly Diagnosed Multiple Myeloma (NDMM) Patients: RV-MM-PI-0752 Phase III Randomized Study. 2018

40. Claudia AM, Stege KN, Mark-David L, Noortje T, Saskia K K, Inge L, Matthijs W, Paula F Y, Yavuz B, Maaike S, de Esther W, Niels W C J v d D, Pieter S, Sonja Zweegman. Efficacy and Tolerability of Ixazomib, Daratumumab and Low Dose Dexamethasone (IDd) in Unfit and Frail Newly Diagnosed Multiple Myeloma (NDMM) Patients; First Interim Safety Analysis of the Phase II HOVON 143 Study. 2018

41. Matino D, Chai-Adisaksopha C, Iorio A. Systematic reviews of prognosis studies: a critical appraisal of five core clinical journals. Diagnostic and Prognostic Research. 2017;1(1):9. doi: 10.1186/s41512-017-0008-z. 10.1186/s41512-017-0008-z

42. Iorio A, Spencer FA, Falavigna M, Alba C, Lang E, Burnand B, et al. Use of GRADE for assessment of evidence about prognosis: rating confidence in estimates of event rates in broad categories of patients. Bmj. 2015;350:h870. doi: 10.1136/bmj.h870. 10.1136/bmj.h870